Re: Information resources

Mikael Nilsson wrote:
>> A lot of the http-range 
>> discussion has been in terms of people, cars, etc; RDF properties are 
>> much closer to the borderline, and so might be a useful category of 
>> thing to help make these definitions sharper.
> 
> I happen to think Properties are non-IR, and I think there are
> compelling arguments for that.
> 
> I think the problem is that people, cars, etc all have an existence
> outside the Web - they are physical things. The class of things you are
> thinking about seem to be abstract entities of different kinds, correct?

Yup, for some sense of "abstract entity" that isn't far away from FRBR Work.

> If you were to say that a Property is an IR, then that same argument
> applies to all abstract entities. Concepts, classes, mathematical
> functions, etc etc.

Well, there are many kinds of abstract entity. I consider RDF terms 
closer to eg. song lyrics or poems, than to pure mathematical things. So 
it might be that some but not all "abstract" things are awww:IR instances.

Classes and properties for RDF can be modelled mathematically using 
sets, model theory, etc. And from within those theories (which change 
over time btw, eg. during the lifetime of the RDFCore and OWL 1.0 groups 
various formalisms were attempted), we'll see things that might say 
"each property is a set" or whatever.

I take the view that dc:creator is a thing with a social history. It 
didn't exist before 1995. We can read about its history in documents 
such as http://www.dlib.org/dlib/january97/oclc/01weibel.html ... and 
the same goes for foaf:mbox_sha1_sum. Once upon a time they didn't 
exist, and now they do. And people say different things about them over 
time, using formal and informal language. Generally (eg. as in RDF) we 
throw away the temporal perspective, since it is complex for humans and 
machines to deal with. So RDF itself treats the world as a static 
timeless whole (although you can of course use it to describe events and 
transitions). From inside that simplified formalism, it looks like RDF 
terms are also timeless. But the Web is bigger than RDF, and will over 
the decades have many machine formalisms. I find it clearer to think of 
RDF vocabularies and their parts as human works, with histories, 
versions and so on just the same.

SKOS might also be an interesting test case here, in that its notion of 
a concept is in some ways performing tasks similar to rdf/owl classes, 
yet it is written in RDF instance vocabulary. Does the AWWW require that 
a SKOS Concept can't be named using a non-#, non-303 http: URI?


> But even so - timbl's distinction between the content of a thing and
> descriptions of that thing apply. That is:  a description is not the
> same as the content of the thing. Or put in other words, what *would* be
> the content of dc:creator? I can't find any message that would convey
> anything other than the definition. 

It might be its authoritative description from the party that published 
it. But I'm not sold on the notion of "its content" anyway; it doesn't 
work well for web services, database lookups, sensor devices attached to 
the Web, etc. It does make sense for "document like objects", but many 
common uses of the Web stretch that metaphor quite a long way; it 
doesn't seem a painful stretch to have it include RDF terms too.

> 
> One potential representation would be all pairs of (subject, object) to
> which the property applies. 

Another would be a description published by the party responsible for 
the term, perhaps RDF,OWL,HTML,RIF, ... perhaps just .txt.

Nothing in the AWWW requires these representations to be lossless 
(roundtripable) copies of "internal" state. Much thinking about the Web 
sorta presumes there's "the real thing" hidden up there inside the Web 
server. That we're talking about computer files, in a directory; which 
itself is a metaphor derrived from the way people file real-world 
documents. But we have to accept that it's a metaphor, and sometimes 
metaphors get over stretched...

Dan

Received on Monday, 26 November 2007 11:48:51 UTC